UPS Runtime Calculator
The runtime printed on a UPS box is measured at full load — which is almost never your load. Add up what you actually plan to plug in, and this works out how many minutes of battery each unit in our database would really give you, plus the minimum capacity you should be shopping for.
Estimates are deliberately conservative. Product links are Amazon affiliate links — purchases support this site at no cost to you.
1. Build Your Load
What are you plugging in?
Typical continuous draw for each device class. If you know your own measured figures, use the custom field at the bottom instead — a plug-in power meter is the most accurate approach.
⚠ Never connect these to a UPS
Laser printers. The fuser draws a 800–1,200W surge every time it warms up. That spike will overload almost any home office UPS and drop everything else connected to it — including the computer you were protecting. Laser printers belong on a surge-only outlet or straight into the wall.
Space heaters, kettles, and anything else that makes heat. A 1,500W heater exceeds the continuous rating of every unit on this page, and there is no scenario where battery-backing one is useful.
Motor loads — shredders, vacuums, mini fridges. Startup surge can be five to eight times the running wattage, and these devices gain nothing from battery backup.
2. What Fits Your Load
| # | Product | Score | Runtime at your load ? |
Load vs capacity ? |
Waveform ? | Rated W | Battery outlets ? |
Verdict |
|---|---|---|---|---|---|---|---|---|
| Select your equipment above to see results. | ||||||||
These rankings come from our full UPS evaluation — 15 units scored across six weighted categories, with the reasoning behind every score published.
Read the full UPS comparisonHow This Calculator Works
The runtime model — and its limits
Battery runtime does not scale linearly with load. Halving the load on a sealed lead-acid battery more than doubles the runtime, because a lower discharge current extracts more usable capacity from the same cells. This is the effect described by Peukert’s law, and ignoring it is why simple “watts divided by capacity” calculators are wrong by a factor of two or more at light loads.
We extrapolate from each unit’s published full-load runtime using a Peukert-style exponent:
The exponent of 1.3 is an empirical fit for the sealed lead-acid batteries used in consumer UPS units. Checked against published manufacturer runtime charts, it lands within about 15% across the 10–50% load range that covers most home offices, and it usually errs low — the published charts are generally a little more generous than this formula.
The added 10 watts accounts for the UPS’s own inverter and control overhead, which is drawn from the battery alongside your equipment. Without it, very light loads produce absurd runtime figures that no real unit achieves.
Three things this model does not capture. Battery age: figures assume a healthy battery, and a three-year-old one can deliver 30–50% less. Temperature: capacity drops meaningfully in a cold garage or a hot closet. Load variability: a desktop PC’s draw swings with what it’s doing, and we use a typical continuous figure rather than a peak.
Treat these as planning estimates, not guarantees. For a load-bearing decision — a home lab, medical equipment, anything where a bad estimate is expensive — size against the manufacturer’s own published runtime chart for the specific model and derate for battery age.
Why we size to 80% of rated capacity
A UPS rated at 900 watts can technically run a 880-watt load. It just shouldn’t. Running close to the rating shortens runtime disproportionately, raises operating temperature, and accelerates battery degradation — and it leaves no headroom for the one day you plug in a second monitor.
Sizing so that your load sits at or below 80% of the unit’s watt rating is the practical convention, and it’s what the “minimum UPS capacity” figure above reflects. It is also worth noting that the watt rating, not the VA rating, is the real limit. A 1500VA unit is typically a 900W unit — a 40% difference that catches people out.
Bottom line: Shop against watts, size for 80%, and treat the VA number as marketing.
Where the device wattages come from
The presets are typical continuous draw figures for each device class, drawn from manufacturer specifications and published power-consumption testing. They are approximations by design — a “mid-range desktop” covers an enormous range of real machines.
If you want precision, a plug-in power meter costs about twenty dollars and measures your actual setup in a few minutes. Enter that figure in the custom field and it replaces the guesswork entirely. This matters most for desktop PCs, which vary more than every other category combined, and least for networking gear, which is tightly clustered.
One deliberate choice: we use typical draw rather than peak. A gaming PC that idles at 120W and peaks at 500W under load will sit near its idle figure during the sort of brief outage a UPS is designed for. Sizing to a peak you’ll never hit during an outage means buying far more UPS than you need.
Bottom line: The presets are good enough for choosing between a 600W and a 900W unit. They are not good enough for shaving the last 10% off a purchase decision — measure if that’s where you are.
Runtime figures are model-derived estimates based on manufacturer specifications researched at time of publication, and are provided for planning purposes only. They are not a substitute for the manufacturer’s published runtime data for a specific model, and should not be relied upon for medical, life-safety, or business-critical continuity planning.